mlsanityData to 7 Oct 2026

Depth-supervised NeRF: Fewer Views and Faster Training for Free

arXiv 2107.02791 ↗First posted 6 Jul 2021cs.CV cs.GR cs.LGcited by 732

Not on the map. Why?

Cited by 55 in Diffusion models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

A commonly observed failure mode of Neural Radiance Field (NeRF) is fitting incorrect geometries when given an insufficient number of input views. One potential reason is that standard volumetric rendering does not enforce the constraint that most of a scene's geometry consist of empty space and opaque surfaces. We formalize the above assumption through DS-NeRF (Depth-supervised Neural Radiance Fields), a loss for learning radiance fields that takes advantage of readily-available depth supervision. We leverage the fact that current NeRF pipelines require images with known camera poses that are typically estimated by running structure-from-motion (SFM). Crucially, SFM also produces sparse 3D points that can be used as "free" depth supervision during training: we add a loss to encourage the distribution of a ray's terminating depth matches a given 3D keypoint, incorporating depth uncertainty. DS-NeRF can render better images given fewer training views while training 2-3x faster. Further, we show that our loss is compatible with other recently proposed NeRF methods, demonstrating that depth is a cheap and easily digestible supervisory signal. And finally, we find that DS-NeRF can support other types of depth supervision such as scanned depth sensors and RGB-D reconstruction outputs.

Key facts. Cited by 732 papers in the map (data through 7 Oct 2026). Builds on 38 of its 49 references that are in the map. First posted on arXiv 6 Jul 2021.

Builds on

38 of 49 references.

References · 38

  1. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  2. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  3. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,682
  4. 2018Occupancy Networks: Learning 3D Reconstruction in Function Spacecited by 2,037
  5. 2020pixelNeRF: Neural Radiance Fields from One or Few Imagescited by 1,574
  6. 2019Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfercited by 1,484
  7. 2020D-NeRF: Neural Radiance Fields for Dynamic Scenescited by 1,459
  8. 2020NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collectionscited by 1,230
  9. 2018Learning Implicit Fields for Generative Shape Modelingcited by 1,060
  10. 2020Neural Sparse Voxel Fieldscited by 959
  11. 2021PlenOctrees for Real-time Rendering of Neural Radiance Fieldscited by 904
  12. 2019Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representationscited by 864
26 more
  1. 2019PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitizationcited by 818
  2. 2020Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenescited by 750
  3. 2021MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereocited by 735
  4. 2018MegaDepth: Learning Single-View Depth Prediction from Internet Photoscited by 731
  5. 2020pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesiscited by 708
  6. 2021IBRNet: Learning Multi-View Image-Based Renderingcited by 698
  7. 2020GRAF: Generative Radiance Fields for 3D-Aware Image Synthesiscited by 688
  8. 2021KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPscited by 672
  9. 2020NeRV: Neural Reflectance and Visibility Fields for Relighting and View Synthesiscited by 493
  10. 2021Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesiscited by 464
  11. 2020Space-time Neural Irradiance Fields for Free-Viewpoint Videocited by 414
  12. 2020INeRF: Inverting Neural Radiance Fields for Pose Estimationcited by 381
  13. 2019SynSin: End-to-end View Synthesis from a Single Imagecited by 364
  14. 20163DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructionscited by 333
  15. 2021Neural RGB-D Surface Reconstructioncited by 312
  16. 2020Single-View View Synthesis with Multiplane Imagescited by 263
  17. 2020Learned Initializations for Optimizing Coordinate-Based Neural Representationscited by 240
  18. 2021DONeRF: Towards Real-Time Rendering of Compact Neural Radiance Fields using Depth Oracle Networkscited by 232
  19. 2021Editing Conditional Radiance Fieldscited by 222
  20. 2020Consistent Video Depth Estimationcited by 220
  21. 20203D Photography using Context-aware Layered Depth Inpaintingcited by 206
  22. 2021Stereo Radiance Fields (SRF): Learning View Synthesis for Sparse Views of Novel Scenescited by 190
  23. 2019Neural Rerendering in the Wildcited by 121
  24. 2020Worldsheet: Wrapping the World in a 3D Sheet for View Synthesis from a Single Imagecited by 77
  25. 2020World-Consistent Video-to-Video Synthesiscited by 69
  26. 2016A Large Dataset of Object Scanscited by 64

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 21; 2022: 98; 2023: 218; 2024: 238; 2025: 110; 2026: 47.

732 papers.

Sort:

By area:

  1. 2021Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstructioncited by 961
  2. 2023Objaverse-XL: A Universe of 10M+ 3D Objectscited by 767
  3. 2023Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstructioncited by 753
  4. 2021RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from Sparse Inputscited by 614
  5. 2021KITTI-360: A Novel Dataset and Benchmarks for Urban Scene Understanding in 2D and 3Dcited by 570
  6. 2024MVSplat: Efficient 3D Gaussian Splatting from Sparse Multi-View Imagescited by 553
  7. 2021Neural Fields in Visual Computing and Beyondcited by 539
  8. 2022MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface Reconstructioncited by 467
  9. 2024Metric3Dv2: A Versatile Monocular Geometric Foundation Model for Zero-shot Metric Depth and Surface Normal Estimationcited by 389
  10. 2024UniDepth: Universal Monocular Metric Depth Estimationcited by 375
  11. 2021Advances in Neural Renderingcited by 365
  12. 2023Compact 3D Gaussian Representation for Radiance Fieldcited by 365
38 more
  1. 2022Fast Dynamic Radiance Fields with Time-Aware Neural Voxelscited by 344
  2. 2021NeRF in the Dark: High Dynamic Range View Synthesis from Noisy Raw Imagescited by 336
  3. 2025GEN3C: 3D-Informed World-Consistent Video Generation with Precise Camera Controlcited by 325
  4. 2023ReconFusion: 3D Reconstruction with Diffusion Priorscited by 317
  5. 2023FreeNeRF: Improving Few-shot Neural Rendering with Free Frequency Regularizationcited by 310
  6. 2023RealFusion: 360deg Reconstruction of Any Object from a Single Imagecited by 305
  7. 2021Dense Depth Priors for Neural Radiance Fields from Sparse Input Viewscited by 301
  8. 2021StyleSDF: High-Resolution 3D-Consistent Image and Geometry Generationcited by 289
  9. 2023FSGS: Real-Time Few-shot View Synthesis using Gaussian Splattingcited by 278
  10. 2021Urban Radiance Fieldscited by 263
  11. 2024DNGaussian: Optimizing Sparse-View 3D Gaussian Radiance Fields with Global-Local Depth Normalizationcited by 256
  12. 2023SparseNeRF: Distilling Depth Ranking for Few-shot Novel View Synthesiscited by 247
  13. 2022NoPe-NeRF: Optimising Neural Radiance Field with No Pose Priorcited by 235
  14. 2024Mini-Splatting: Representing Scenes with a Constrained Number of Gaussianscited by 208
  15. 2024GaussianPro: 3D Gaussian Splatting with Progressive Propagationcited by 199
  16. 2021GeoNeRF: Generalizing NeRF with Geometry Priorscited by 195
  17. 2022SPARF: Neural Radiance Fields from Sparse and Noisy Posescited by 195
  18. 2022NeRF-SLAM: Real-Time Dense Monocular SLAM with Neural Radiance Fieldscited by 188
  19. 2021InfoNeRF: Ray Entropy Minimization for Few-Shot Neural Volume Renderingcited by 180
  20. 2023Depth-Regularized Optimization for 3D Gaussian Splatting in Few-Shot Imagescited by 172
  21. 2023MARS: An Instance-aware, Modular and Realistic Simulator for Autonomous Drivingcited by 170
  22. 2025UniDepthV2: Universal Monocular Metric Depth Estimation Made Simplercited by 170
  23. 2023GauHuman: Articulated Gaussian Splatting from Monocular Human Videoscited by 168
  24. 2022SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Imagecited by 165
  25. 2023PF-LRM: Pose-Free Large Reconstruction Model for Joint Pose and Shape Predictioncited by 165
  26. 2024CamCo: Camera-Controllable 3D-Consistent Image-to-Video Generationcited by 163
  27. 2025Difix3D+: Improving 3D Reconstructions with Single-Step Diffusion Modelscited by 160
  28. 2022ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of Signed Distance Fieldscited by 158
  29. 2022Instant Volumetric Head Avatarscited by 157
  30. 2022NeRF: Neural Radiance Field in 3D Vision: A Comprehensive Review (Updated Post-Gaussian Splatting)cited by 156
  31. 2022NeRDi: Single-View NeRF Synthesis with Language-Guided Diffusion as General Image Priorscited by 156
  32. 2022Panoptic NeRF: 3D-to-2D Label Transfer for Panoptic Urban Scene Segmentationcited by 152
  33. 2023SUDS: Scalable Urban Dynamic Scenescited by 151
  34. 2021Efficient Neural Radiance Fields for Interactive Free-viewpoint Videocited by 150
  35. 2022NeuralLift-360: Lifting An In-the-wild 2D Photo to A 3D Object with 360deg Viewscited by 148
  36. 2022SPIn-NeRF: Multiview Segmentation and Perceptual Inpainting with Neural Radiance Fieldscited by 144
  37. 2023Relightable 3D Gaussians: Realistic Point Cloud Relighting with BRDF Decomposition and Ray Tracingcited by 138
  38. 2022D2NeRF: Self-Supervised Decoupling of Dynamic and Static Objects from a Monocular Videocited by 136

50 most cited of 732 All 732 (JSON)